Natural Climatic Cycles

4.2 Natural Climatic Cycles

Climate is defined as an aggregate of weather conditions, representing a general pattern of weather variations at a location or in a region. It includes average weather conditions, as well as the variability of elements and information on the occurrence of extreme events (Lutgens and Tarbuck, 1995). The nature of both weather and climate is expressed in terms of basic elements, the most important of which are (1) the temperature of the air, (2) the humidity of the air, (3) the type and amount of cloudiness, (4) the type and amount of precipitation, (5) the pressure exerted by the air, and (6) the speed and direction of the wind. These elements constitute the variables by which weather patterns and climatic types are depicted (Lutgens and Tarbuck, 1995). The main difference between weather and climate is the time scale at which these basic elements change. Weather is constantly changing, sometimes from hour to hour, and these changes create almost an infinite variety of weather conditions at any given time and place. In comparison, climate changes

ClimateChange

are much more subtle and were, until relatively recently, considered important for time scales of hundreds of years or more, and usually only discussed in academic circles.

A more broad definition of climate is that it represents the long-term behavior of the interactive climate system, which consists of the atmosphere, hydrosphere, lithosphere, biosphere, and cryosphere or ice and snow that are accumulated on the earth’s surface. To understand fully and to predict changes in the atmosphere component of the climate system, one must understand the sun, oceans, ice sheets, solid earth, and all forms of life (Lutgens and Tarbuck, 1995).

The most significant theory relating earth motions and long-term climate change, later confirmed with geologic and paleoclimatic evidence collected from around the globe, was developed in the 1930s by the Serbian mathematician and astrophysicist Milutin Milankovitch, professor at the University of Belgrade. His work titled Kanon der Erdbestrahlung und seine Anwendung auf das Eiszeitenproblem (Canon of Insolation of the Earth and Its Application to the Problem of the Ice Ages) was published in German in 1941 by the Royal Serbian Academy, but was largely ignored by the international scientific community. In 1969, it was translated into English and published with the title Canon of Insolation of the Ice-Age Problem by the U.S. Department of Commerce and the National Science Foundation, Washington, DC. In 1976, a study published in the journal Science examined deep-sea sediment cores and found that Milankovitch’s theory did in fact correspond to periods of climate change (Hays et al., 1976). Specifically, the authors were able to analyze the record of temperature change going back 450,000 years and found that major variations in climate were closely associated with changes in the geometry (eccentricity, obliquity, and precession) of the earth’s orbit; ice ages had indeed occurred when the earth was going through different stages of orbital variation. Since this study, the National Research Council of the U.S. National Academy of Sciences has embraced the Milankovitch cycle model (NRC, 1982):

... orbital variations remain the most thoroughly examined mechanism of climatic change on time scales of tens of thousands of years and are by far the clearest case of a direct effect of changing insolation on the lower atmosphere of Earth.

Milankovitch was intrigued by the puzzle of climate change and studied climate records, noting differences over time. He theorized that global climate change was brought about by regular changes in earth’s axis, tilt, and orbit that altered the planet’s re- lationship to the sun, triggering ice ages. Milankovitch determined that the earth wobbles in its orbit and calculated the slow changes in the earth’s orbit by careful measurement of the position of the stars and using the gravitational pull of other planets and stars. The three variables quantified by Milankovitch are now known as Milankovitch cycles:

1. Eccentricity cycle of the earth’s orbit; every 90,000 to 100,000 years there is a change in the earth’s orbit about the sun. Its almost circular orbit becomes more elliptical, taking earth farther from the sun.

2. The tilt of the earth’s axis or obliquity cycle; on average, every 40,000 years there is a change in the tilt of the earth’s equatorial plane in relation to its orbital plane, moving either the northern or the southern hemisphere farther from the sun.

3. Precession or orientation of the earth’s rotational axis; on average, every 22,000 years there is a slight change in its wobble (the earth does not rotate perfectly like a wheel about an axis; it spins like a wobbling top).

298 ChapterFour

.06 ion .04

-.02 ial precess -.04 ax -.06

24.0 ity 23.5 iqu 23.0 obl

future past

Time (thousands of years)

F IGURE 4.2 Calculated values for 300,000 years of Milankovitch cycles. (From NASA, 2007; source: Berger and Loutre, 1991.)

These cycles mean that during certain periods there is less solar energy arriving to the earth, resulting in less melting of snow and ice. Instead of melting, these cold expanses of frozen water grow. The snow and ice last longer and, over many seasons, begin to accumulate. Snow and ice reflect some sunlight back into space, which also contributes to cooling. Temperatures drop, and glaciers begin to advance (Tesla Memorial Society of New York, 2007).

The climate is influenced by all three cycles that can combine in a number of different ways, sometimes strongly reinforcing each other and sometimes working against each other. The general influence of the Milankovitch cycles on the long-term climate and their current state is presented below based on NASA (2007; see also Fig. 4.2).

The eccentricity of the earth’s orbit changes slowly over time from nearly zero (circu- lar) to 0.07 (eccentric). As the orbit becomes more eccentric (oval), the difference between the distance from the sun to the earth at perihelion (closest approach) and aphelion (furthest away) becomes greater and greater. Currently, a difference of only 3 percent (5 million km) exists between perihelion, which occurs on or about January 3, and aphelion, which occurs on or about July 4. This difference in distance amounts to about a 6 percent increase in incoming solar radiation (insolation) from July to January. The current trend of eccentricity is decreasing. When the orbit is highly elliptical, the amount of insolation received at perihelion would be on the order of 20 to 30 percent greater than at aphelion, resulting in a substantially different climate from what we experience today. ◦

Today, the earth’s axis is tilted 23.5 from the plane of its orbit around the sun. During a

cycle that averages about 40,000 years, the tilt of the axis varies between 22.1 ◦ and 24.5 . Because of tilt changes, the seasons as we know them can become exaggerated. More

tilt means more severe seasons—warmer summers and colder winters; less tilt means

ClimateChange

less severe seasons—cooler summers and milder winters. It is the cool summers that allow snow and ice to last from year to year in high latitudes, eventually building up into massive ice sheets. An earth covered with more snow reflects more of the sun’s energy into space, causing additional cooling. The current trend in the earth’s axis tilt is decreasing.

Changes in axial precession alter the dates of perihelion and aphelion and, therefore, increase the seasonal contrast in one hemisphere and decrease the seasonal contrast in the other hemisphere. If a hemisphere is pointed toward the sun at perihelion, that hemisphere will be pointing away at aphelion and the difference in seasons will be more extreme. This seasonal effect is reversed for the opposite hemisphere. Currently, the northern hemisphere summer occurs near aphelion, which means that the northern hemisphere should have somewhat less extremes between the seasons. The climatic precession is close to its peak and shows a decreasing trend.

Although the Milankovitch cycles can explain long-term climatic changes on geo- logic time scales (on the order of tens of thousands of years or more), their long duration makes them ineffective tools to explain or predict changes that are of significance for water resources evaluation and planning, namely, at time scales of decades to centuries. How- ever, what we can learn from the well-established science of long-term climate change and the geologic evidence of it occurring in the past is that it will inevitably occur in the future as well. A fourth cycle, not addressed by Milankovitch, may accelerate natural climate change—human activity on earth. The photograph in Fig. 4.3 is an evidence that

F IGURE 4.3 Cave divers in submerged cave passages with an abundance of speleothems— stalactites, stalagmites, flowstone, and columns—formed prior to submergence. Nohoch Nah

Chich in the Yucatan Peninsula, 2007. (Photograph courtesy of David Rhea, Global Underwater Explorers.)

300 ChapterFour

the sea level in the past was lower than it is today. One of the reasons is that the ice accumulated on the continents during the last ice age did melt to a large degree, causing

a significant global sea level rise. As a result, the water table in coastal aquifers also rose, as evidenced from submerged caves in karst regions such as the Yucatan Peninsula in Mexico. Speleothems visible in the photograph could have only been formed when the cave was not submerged. Vast cave systems, many of which are now completely filled with freshwater like this one, were developed in the Yucatan when the sea level was lower than today.

Accurate and systematic measurements of weather and climate elements are paramount to fully understanding the climate of a region and anticipating future cli- matic changes that may impact water supplies. Unfortunately, records of air temperature and precipitation, the most important direct measures of climate, go back only several hundred years in Europe and less than that in other parts of the world. The situation is even worse with hydrologic measurements of streamflows or spring flows and worse yet with records of groundwater levels, the two most important direct measures of freshwa- ter budget. Even though the time record of direct climatic and hydrologic measurements is increasing, it is becoming more and more evident that 100 hundred years or so is still too short to capture the statistics necessary for a more accurate probability analysis of the extreme climate events such as floods and droughts. For example, it was during a wet period in the measured hydrologic record that the 1922 Colorado River Compact estab- lished the basic apportionment of the river between the Upper and Lower Colorado River Basins in the United States. At the time of Compact negotiations, it was thought that an

average annual flow volume of about 21 million acre-ft (MAF; 1 acre-ft equals 136.8 m 3 ) was available for apportionment. The Compact provided for 7.5 MAF of consumptive use annually for each of the basins, plus the right for the lower basin to develop 1 MAF of consumptive use annually. Subsequently, a 1944 Treaty with Mexico provided a volume of water of 1.5 MAF annually for Mexico. During the period of measured hydrology now available, the river’s average annual natural flow has been about 15 MAF at Lee Ferry (ACWA and CRWUAC, 2005). This over-allocation of the Colorado River is now causing many political and societal problems in the region.

Studies in the last two decades have revealed that some climatic fluctuations once thought to be local phenomena are part of a large-scale atmospheric circulation that periodically affects global weather and contributes to long-term climate characteristics of different world regions. The best known and the most studied is ENSO (El Ni ˜no–Southern Oscillation). Centuries ago, the local residents on the coasts of Ecuador and Peru named

a regular annual weather event El Ni ˜no (“the child”) after the Christ child because it usually appeared during the Christmas season. During this event that lasts a few weeks,

a weak, warm countercurrent flows southward along the coasts of Ecuador and Peru, replacing the cold Peruvian current. However, every 3 to 7 years this countercurrent is unusually warm and strong and is accompanied by a pool of warm ocean surface water in the central and eastern Pacific, which influences weather worldwide (Lutgens and Tarbuck, 1995).

The second strongest El Ni ˜no on record occurred in 1982 and 1983 (Fig. 4.4) and was blamed for weather extremes of a variety of types in many parts of the world. Heavy rains and flooding affected normally dry portions of Peru and Ecuador. Australia, Indonesia, and the Philippines experienced severe droughts, while one of the warmest winters on record was followed by one of the wettest springs for much of the United States. Heavy snows in the Sierra Nevada and the mountains of Utah and Colorado led to mudflows

ClimateChange

F IGURE 4.4 Oceanic Ni˜ no Index, ONI ( ◦ C), evolution since 1950. ONI is the principal measure for monitoring, assessing, and predicting ENSO. Positive values greater than +0.5 generally indicate El Ni˜ no conditions, and negative values less than –0.5 generally indicate La Ni˜ na conditions. (From CPC/NCEP, 2007b.)

and flooding in Utah and Nevada and along the Colorado River in the spring of 1983. The unusual rains brought floods to the Gulf States and Cuba. Unfortunately, as discussed by Lutgens and Tarbuck (1995), the effects of El Ni ˜no are highly variable, depending in part on the temperatures and size of the warm pools in the Pacific. During one El Ni ˜no, an area may experience flooding, only to be hit by drought during the next event. It is such extreme events that water managers both fear and are constantly preparing for. Climate Prediction Center (CPC) of the National Weather Service, National Oceanic and Aeronautic Administration (NOAA), maintains a Web page dedicated to the research and weather predictions associated with El Ni ˜no and La Ni ˜na events (CPC, 2007a).

The probability of floods and droughts is the key design element for water supply systems relying on surface water. Although systems based on groundwater are much less vulnerable to extreme weather events, they too can be stressed during prolonged droughts as a result of increased demand for water. Edwards and Redmond’s discussion on the climatic conditions in the Colorado River Basin, the United States, illustrates the

302 ChapterFour

importance of understanding and predicting cyclical climate patterns for water supply management (Edwards and Redmond, 2005).

The waters of the Colorado River originate primarily in the high mountain basins of Colorado, Utah, and Wyoming and flow through seven states and two countries. With headwaters about 1500 and 1700 mi from the Gulf of California, the Colorado and the Green Rivers, respectively, contribute equally to about 80 percent of the flow into Lake Powell, with the remainder mostly from the San Juan River and Mountain Range. Within

the 242,000 mi 2 U.S. portion of the Colorado River Basin (Basin), the highest one-seventh of the basin supplies about six-sevenths of the total flow, and many of the lower river reaches lose water under natural conditions. Most of the precipitation supply falls in winter as snow on interior mountain ranges. Spring precipitation can be important, but summer precipitation is usually nearly negligible in altering water supply, although it does influence demand. Thus, climatic influences on the interior mountain ranges are key factors governing the supply of water in the river from 1 year or decade to the next.

The warm phase of ENSO, El Ni ˜no, typically brings wet and cool winters to the southwest United States and dry and warm winters to the Pacific northwest and northern Rockies. Overall, El Ni ˜no winters tend to have more wet days, more precipitation per wet day, and more persistent wet episodes in the southwestern United States. All of these favor increased runoff. Notably, extremely high or low flow is better correlated with ENSO than is total runoff volume. The opposite cool phase of ENSO, La Ni ˜na, has been reliably associated with dry and warm winters in the Southwest for the past

75 years and, to a less reliable extent, with wet and cool winters in the northern West. The understanding of ENSO and its effects on the Basin are crucial in predicting winter snowpack. So far, western North America climate relationships to ENSO appear to be confined to the winter, with slight or ambiguous associations with summer climate. In the Colorado River Basin, the strongest relationships are seen in the lower basin, south of the San Juan Mountains of Colorado. The relationship becomes less clear farther north and begins to have the opposite effect in the upper Green River Basin and the Wind River Mountains in Wyoming.

On the basis of the analysis of the Colorado River Basin’s climate, Edwards and Redmond (2005) offer the following summary:

Through multiple re-use, the river provides water supply needs for 28 million people, a number expected to continue to grow in coming decades as the Southwest’s population continues to expand at the fastest rate in the nation. The river basin has been developed through an extensive infrastructure system that was designed to buffer against the region’s significant climate variability. Of note, however, the system has not been thoroughly tested by events of the magnitude that we have learned from the paleoclimate record may occur. The recent drought has provided a taste of what is possible, though not the full meal.

Figures 4.5. and 4.6 illustrate the combined impact of several recent droughts and water use on Lake Mead, one of the most important water resources in the American west. Created in the 1930s, it ensures a steady water supply for Arizona, Nevada, California, and northern Mexico by holding back the flow of the Colorado River behind the Hoover dam. It is one of the largest water reservoirs in the world. When full, the lake contains roughly the same amount of water as would have otherwise flowed through the Colorado River over a 2-year period: roughly 36 trillion liters (9.3 trillion gallons). Ninety percent of southern Nevada’s water comes from Lake Mead, with releases being regulated by the

ClimateChange

F IGURE 4.5 Lake Mead in 2004. (Photograph courtesy of Andy Pernick, the U.S. Bureau of Conservation.)

Lake Mead elevation (ft) 1,000

Year F IGURE 4.6 Lake Mead level for September, 1935–2007. (Source of raw data: U.S. Bureau of

304 ChapterFour

Southern Nevada Water Authority. When the water levels in the lake are declining and expected to cross below an elevation of 1145 ft, the Water Authority declares a drought watch. Once the water is below 1145 ft, the watch is shifted to a drought alert. If the lake level drops below 1125 ft, a drought emergency goes into effect. Each of the water-level alert states triggers various water restrictions and practices in the area, from restrictions on watering gardens, washing cars, running fountains in civic parks, and public places to increases in the rates charged for water to encourage conservation (Allen, 2003).

In April of 2007, the water level dropped below 1125 ft for the first time since 1965 and remained below this benchmark through September 2007, the last month with data available to the author. The graph in Fig. 4.6 shows that it took about 20 years for the lake levels to recover from the 1965 low. As discussed throughout this chapter, the over- allocation of the Colorado River water, combined with population growth and the impact of droughts, is putting additional stress on groundwater resources in this semiarid to arid region where natural groundwater recharge is very low.

4.2.1 Droughts

As pointed out by the National Drought Mitigation Center (NDMC, 2007), drought is a normal, recurrent feature of climate, although many erroneously consider it a rare and random event. Graphs like the one shown in Fig. 4.7 remind us of this simple fact. It is understandable, however, that every current drought may always be the hardest ever for the people affected by it, since human memory tends to block unpleasant experiences from the past. (Note: As opposed to the general public, water resources managers are not expected to have this characteristic.) When droughts are of historic proportions, they may trigger major societal changes and forever impact the use and management of water resources. For example, the major drought of the twentieth century in the United States, in terms of duration and spatial extent, is considered to be the 1930s Dust Bowl drought, which lasted up to 7 years in some areas of the Great Plains (Fig. 4.8). This drought, memorialized in John Steinbeck’s novel The Grapes of Wrath was so severe, widespread, and lengthy that it resulted in a mass migration of millions of people from the Great Plains to the western United States in search of jobs and better living conditions. It also dramatically changed agricultural practices including the unprecedented large-scale use of groundwater for irrigation across the Great Plains and throughout the American west.

Although drought has scores of definitions, it originates from a deficiency of precipi- tation over an extended period of time, usually a season or more. This deficiency results in a water shortage for some activity, group, or environmental sector. Drought should be considered relative to some long-term average condition of balance between precipita- tion and evapotranspiration in a particular area, a condition often perceived as “normal.”

1912- 1918- 1922- 1929-34

1959- 1978- 1987-

2000 F IGURE 4.7 California’s multiyear historical dry periods of statewide or major regional extent,

1850–2000. Dry periods prior to 1900 are estimated from limited data. (Source: http://watersupplyconditions.water.ca.gov/.)

ClimateChange

F IGURE 4.8 A great “roller” of dust moves across the land in Colorado during the Dust Bowl of 1930s. (Photograph courtesy of National Resources Conservation Service.)

It is also related to the timing (such as principal season of occurrence, delays in the start of the rainy season, and occurrence of rains in relation to principal crop growth stages) and the effectiveness of the rains such as rainfall intensity and number of rainfall events. Other climatic factors such as high temperature, high wind, and low relative humidity are often associated with droughts in many regions of the world and can significantly aggravate their severity (NDMC, 2007).

Drought should not be viewed as merely a physical phenomenon or natural event. Its impacts on society result from the interplay between a natural event (less precipitation than expected resulting from natural climatic variability) and the demand people place on water supply. Human beings often exacerbate the impact of drought. Recent droughts in both developing and developed countries and the resulting economic and environmental impacts and personal hardships have underscored the vulnerability of all societies to this natural hazard (NDMC, 2007).

Two main drought definitions are conceptual and operational. Conceptual defini- tions, formulated in broad terms, help the general public understand the concept of drought. For example, “drought is a protracted period of deficient precipitation result- ing in extensive damage to crops, resulting in loss of yield.” Conceptual definitions may also be important in establishing drought policy. For example, Australian drought policy incorporates an understanding of normal climate variability into its definition of drought. The country provides financial assistance to farmers only under “excep- tional drought circumstances,” when drought conditions are beyond those that could be

306 ChapterFour

considered part of normal risk management. Declarations of exceptional drought are based on science-driven assessments. Previously, when drought was less well defined from a policy standpoint and less well understood by farmers, some farmers in the semi- arid Australian climate claimed drought assistance every few years (NDMC, 2007).

Operational definitions of drought help identify the beginning, end, and degree of severity of a drought. To determine the beginning of drought, operational definitions specify the degree of departure from the average of precipitation or some other climatic variable over some time period. This is usually done by comparing the current situation to the historical average, often based on a 30-year period of record. The threshold identified as the beginning of a drought (e.g., 75 percent of average precipitation over a specified time period) is usually established somewhat arbitrarily, rather than on the basis of its precise relationship to specific impacts.

An operational definition for agriculture might compare daily precipitation values to evapotranspiration rates to determine the rate of soil moisture depletion and then express these relationships in terms of drought effects on plant behavior (i.e., growth and yield) at various stages of crop development. Operational definitions can also be used to analyze drought frequency, severity, and duration for a given historical period. Developing a climatology of drought for a region provides a greater understanding of its characteristics and the probability of recurrence at various levels of severity. Information of this type is extremely beneficial in the development of response and mitigation strategies and preparedness plans (NDMC, 2007).

Although the major droughts of the twentieth century, the 1930s Dust Bowl and the 1950s droughts, had the most severe impact on the central United States., droughts regularly occur all across North America. Florida suffered from the 1998 drought along with the states of Oklahoma and Texas. Extensive drought-induced fires burned over 475,000 acres in Florida and cost $500 million in damages. In the same year, Canada suffered its fifth-highest fire occurrence season in 25 years. Starting in 1998, 3 years of record low rainfall plagued northern Mexico. The year 1998 was declared the worst drought in 70 years. It became worse as 1999 spring rainfalls were 93 percent below normal. The government of Mexico declared five northern states as disaster zones in 1999 and nine in 2000. The U.S. west coast experienced a 6-year drought in the late 1980s and early 1990s, causing Californians to take aggressive water conservation measures. Even the typically humid northeastern United States experienced a 5-year drought in the 1960s, draining reservoirs in New York City down to 25 percent of capacity (NCDC, 2007a).

The impact of the 2007 drought in southern California and the southeastern United States is yet to be assessed, although it is already apparent that it will have a major influence on water management decisions. For example, the governor of California and the democratic party-led state legislature came to an impasse over the emergency state investments in major water supply projects, with the governor favoring construction of large surface water reservoirs and the legislature favoring the use of groundwater and artificial aquifer recharge.

Drought is a natural hazard that cumulatively has affected more people in North America than any other natural hazard (Riebsame et al., 1991). In the United States, the cost of losses due to drought averages $6 to $8 billion every year but range as high as $39 billion for the 3-year drought of 1987 to 1989, which was the most costly natural disaster documented in U.S. history at the time. Continuing uncertainty in drought prediction contributes to crop insurance payouts of over $175 million per year in western Canada (NCDC, 2007a).

ClimateChange

Natural Climate Variability

Precipitation deficiency

High temp., high winds, low

(amount, intensity, timing)

ical Reduced infiltration, runoff,

relative humidity, greater sunshine, less cloud cover

deep percolation, and

Drought groundwater recharge

Increased evaporation

Meteorolog ion)

and transpiration

t (durat

Soil water deficiency

icultural ime

Plant water stress, reduced

Agr T Drough

biomass and yield

Reduced streamflow, inflow to ical t reservoirs, lakes, and ponds;

reduced wetlands.

wildlife habitat

Drough Hydrolog

Economic impacts

Social Impacts

Environmental Impacts

F IGURE 4.9 Three categories of drought identified by the National Drought Mitigation Center. (From NDMC, 2007.)

Figure 4.9 illustrates the concept of three different drought categories identified by the National Drought Mitigation Center as follows:

Agricultural drought links various characteristics of meteorological (or hydrological) drought to agricultural impacts, focusing on precipitation shortages, differences be- tween actual and potential evapotranspiration, soil water deficits, reduced ground- water or reservoir levels, and so forth. Crop water demand depends on prevailing weather conditions, biological characteristics of the specific plant, its stage of growth, and the physical and biological properties of the soil. A good definition of agricul- tural drought should be able to account for the variable susceptibility of crops during different stages of crop development, from emergence to maturity. Deficient topsoil moisture at planting may hinder germination, leading to low plant populations per hectare and a reduction of final yield. However, if topsoil moisture is sufficient for early growth requirements, deficiencies in subsoil moisture at this early stage may not affect final yield if subsoil moisture is replenished as the growing season progresses or if rainfall meets plant water needs.

Hydrological drought is concerned with the effects of periods of precipitation (including snowfall) shortfalls on surface or subsurface water supply (such as streamflow, reser- voir and lake levels, and groundwater). The frequency and severity of hydrological drought are often defined on a watershed or river basin scale. Although all droughts originate with a deficiency of precipitation, hydrologists are more concerned with how

308 ChapterFour

this deficiency plays out through the hydrologic system. Hydrological droughts are usually out of phase with or lag the occurrence of meteorological and agricultural droughts. It takes longer for precipitation deficiencies to show up in components of the hydrological system such as soil moisture, streamflow, and groundwater and reser- voir levels. As a result, these impacts are out of phase with impacts in other economic sectors. For example, precipitation deficiency may result in rapid depletion of soil moisture that is almost immediately discernible to agriculturalists, but the impact of this deficiency on reservoir levels may not affect hydroelectric power production or recreational uses for many months. Also, water in hydrologic storage systems (e.g., reservoirs and rivers) is often used for multiple and competing purposes (e.g., flood control, irrigation, recreation, navigation, hydropower, and wildlife habitat), further complicating the sequence and quantification of impacts. Competition for water in these storage systems escalates during drought, and conflicts between water users increase significantly.

Socioeconomic drought definition associates the supply and demand of some economic good with elements of meteorological, hydrological, and agricultural drought. It dif- fers from the aforementioned types of drought because its occurrence depends on the time and space processes of supply and demand to identify or classify droughts. The supply of many economic goods, such as water, forage, food grains, fish, and hydro- electric power, depends on weather. Because of the natural variability of climate, water supply is ample in some years but unable to meet human and environmental needs in other years. Socioeconomic drought occurs when the demand for an economic good exceeds supply as a result of a weather-related shortfall in water supply. For example, in Uruguay in 1988 to 1989, drought resulted in significantly reduced hydroelectric power production because power plants were dependent on streamflow rather than storage for power generation. Reducing hydroelectric power production required the government to convert to more expensive (imported) petroleum and enforce stringent energy conservation measures to meet the nation’s power needs.

In most instances, the demand for economic goods is increasing as a result of in- creasing population and per capita consumption. Supply may also increase because of improved production efficiency, technology, or the construction of reservoirs that in- crease surface water storage capacity. If both supply and demand are increasing, the critical factor is the relative rate of change. Is demand increasing more rapidly than sup- ply? If so, vulnerability and the incidence of drought may increase in the future as supply and demand trends converge (NDMC, 2007).

The sequence of impacts associated with meteorological, agricultural, and hydrolog- ical drought further emphasizes their differences. When drought begins, the agricultural sector is usually the first to be affected because of its heavy dependence on stored soil water. Soil water can be rapidly depleted during extended dry periods. If precipitation deficiencies continue, then people dependent on other sources of water will begin to feel the effects of the shortage. Those who rely on surface water (i.e., reservoirs and lakes) and groundwater are usually the last to be affected. A short-term drought that persists for 3 to 6 months may have little impact on these sectors, depending on the characteristics of the hydrologic system and water use requirements.

When precipitation returns to normal and meteorological drought conditions have abated, the sequence is repeated for the recovery of surface and subsurface water supplies. Soil water reserves are replenished first, followed by streamflow, reservoirs

ClimateChange

and lakes, and finally groundwater. Drought impacts may diminish rapidly in the agri- cultural sector because of its reliance on soil water, but linger for months or even years in other sectors dependent on stored surface or subsurface supplies. Groundwater users, often the last to be affected by drought during its onset, may be the last to experience

a return to normal water levels. The length of the recovery period is a function of the intensity of the drought, its duration, and the quantity of precipitation received as the episode terminates (NDMC, 2007).

Drought Indices Drought indices assimilate data on rainfall, snowpack, streamflow, and other water sup- ply indicators into a comprehensible big picture. A drought index value is typically a single number, far more useful than raw data for decision making. There are several indices that measure how much precipitation for a given period of time has deviated from historically established norms. Although none of the major indices is inherently superior to the rest in all circumstances, some indices are better suited than others for certain uses. For example, the Palmer Drought Severity Index (PDSI or The Palmer) has been widely used by the U.S. Department of Agriculture to determine when to grant emergency drought assistance. The Palmer index is better when working with large ar- eas of uniform topography. Western states, with mountainous terrain and the resulting complex regional microclimates, find it useful to supplement Palmer values with other indices such as the Surface Water Supply Index, which takes snowpack and other unique conditions into account. Detailed discussion of various drought indices, including their advantages and drawbacks, is given by Hayes (2007).

The National Drought Mitigation Center is using a newer index, the Standardized Precipitation Index (SPI), to monitor moisture supply conditions. Some distinguishing traits of this index are that it identifies emerging drought months sooner than the Palmer index and that it is computed on various time scales. The understanding that a deficit of precipitation has different impacts on groundwater, reservoir storage, soil moisture, snowpack, and streamflow led McKee, Doesken, and Kleist of Colorado State University to develop the SPI in 1993. The SPI was designed to quantify the precipitation deficit for multiple time scales. These time scales reflect the impact of drought on the availability of different water resources. Soil moisture conditions respond to precipitation anomalies on a relatively short scale. Groundwater, streamflow, and reservoir storage reflect the longer-term precipitation anomalies. For these reasons, McKee et al. (1993) originally calculated the SPI for 3-, 6-, 12-, 24-, and 48-month time scales.

The SPI calculation for any location is based on the long-term precipitation record for a desired period. This long-term record is fitted to a probability distribution, which is then transformed into a normal distribution so that the mean SPI for the location and desired period is zero (Edwards and McKee, 1997). Positive SPI values indicate greater than median precipitation, and negative values indicate less than median precipitation. Because the SPI is normalized, wetter and drier climates can be represented in the same way, and wet periods can also be monitored using the SPI. PSI values smaller than –2 indicate extreme drought and greater than 2 extreme wet conditions.

Occurrence of Droughts Instrumental records of drought for the United States extend back approximately 100 years. These records capture the twentieth-century droughts but are too short to as- sess the reoccurrence of major droughts such as those of the 1930s, 1950s, and 2000s.

310 ChapterFour

As droughts continue to have increasingly costly impacts on the society, economy, and environment, it is becoming even more important to put the severe droughts of the twen- tieth and the beginning of the twenty-first century into a long-term water management perspective. This perspective can be gained through the use of paleoclimatic records of drought, called proxies. Paleoclimate is the climate of the past, before the development of weather-recording instruments, and is documented in biological and geological systems that reflect variations in climate in their structure. Different proxies record variations in drought conditions on the order of single seasons to decadal- and century-scale changes, providing scientists with the information about both rapid and slow changes and short and long periods of drought.

Historical records, such as diaries and newspaper accounts, can provide detailed information about droughts for the last 200 (mid-western and western United States) or 300 (eastern United States) years. Tree-ring records can extend back 300 years in most areas and thousands of years in some regions. In trees that are sensitive to drought conditions, tree rings provide a record of drought for each year of the tree’s growth. Geologic evidence is used for records longer than those provided by trees and historical accounts and for regions where such accounts are absent. It includes analysis of lake sediments and their paleontologic content, and sand dunes (NCDC, 2007a).

Lake sediments, if the cores of the sediments are sampled at very frequent intervals, can provide information about variations occurring at frequencies less than a decade in length. Fluctuations in lake level can be recorded from beach material sediments (geologic bath tub rings), which are deposited either high (further from the center under wetter conditions) or low (closer to the center under drier conditions) within a basin as the water depth and thus lake level change in response to drought. Droughts can increase the salinity of lakes, changing the species of small, lake-dwelling organisms that occur within a lake. Pollen grains get washed or blown into lakes and accumulate in sediments. Different types of pollen in lake sediments reflect the vegetation around the lake and the climate conditions that are favorable for that vegetation. For example, a change in the type of pollen found in sediments from an abundance of grass pollen to an abundance of sage pollen can indicate a change from wet to dry conditions (NCDC, 2007a).

Records of more extreme environmental changes can be found by investigating the layers within sand dunes. The sand layers are interspersed among layers of soil material produced under more wet conditions, between the times when the sand dune was active. For a soil layer to develop, the climate needs to be wet for an extended period of time; such layers therefore reflect slower, longer-lasting changes.

Large areas of the intermountain basins of the western United States contain sand dunes and other dune-related features, most of which are now stabilized by vegetation. Sand dunes and sand sheets were deposited by the wind in times of drought and contain

a wealth of information about episodes of drought and aridity over the course of the Holocene, which is the period since the end of the most recent widespread glaciation, about 10,000 years ago. The soil layers, which are interspersed with sand layers, con- tain organic materials and can be dated with radiocarbon dating techniques. The dates from the soil layers between layers of sand can be used to bracket times of drought as signified by the presence of sand. Since there is a lag in time in the vegetation and dune response to climate conditions, this record is fairly coarse in terms of time scales that it can resolve (typically centuries or longer). In addition, radiocarbon dating, with a dating precision of 5 percent (or more during certain periods in the Holocene), contributes to low temporal resolution of this record. However, recent work has used optically stimulated

ClimateChange

200 Streamflow (1000 acre ft)

Year F IGURE 4.10 Calibration/verification period (1916–2002) for a tree-ring reconstruction of the

South Platte River at South Platte, CO. The reconstruction explains most of the variance (R 2 = 0.76) of the gaged record and captures the extreme low flows, including 2002. The gaged flow record, corrected for depletions, was provided by Denver Water. Units for flow are 1000 acre-ft. (From Woodhouse and Lukas, 2005.)

luminescence techniques to date sand grains, producing records with a decadal scale resolution for the past 1000 years (Woodhouse, 2005).

Tree rings provide annually or seasonally resolved data that are precisely dated to the calendar year. Tree-ring records commonly extend 300 to 500 years into the past, and

a small number are thousands of years long. Trees that are sensitive to climate reflect variations in climate in the width of their annual rings. Thus, the ring-width patterns contain records of past climate. Trees that grow in arid or semiarid areas and on open, dry, south-facing slopes are stressed by a lack of moisture. These trees can be used for reconstructing climate variables such as precipitation, streamflow, and drought. To de- velop a reconstruction of past climate, tree-ring data are calibrated with an instrumental record for the period of years common to both. This process yields a statistical model that is applied to the full length of the tree-ring data to generate a reconstruction of past climate. The reconstructions are only estimates of past climate, as the tree-ring-based re- constructions do not explain all the variance in the instrumental records. However, they can explain up to 60 to 75 percent of the total variance in an instrumental record (Fig. 4.10; Woodhouse and Lukas, 2005). A detailed explanation of the tree-ring paleoclimate dating including examples is given in Meko et al. (1991) and Cook et al. (1999).

A remarkably widespread and persistent period of drought in the late-sixteenth cen- tury is evident in a large number of various proxy records for the western United States. Tree-ring data document drought conditions that ranged across western North America from northern Mexico to British Columbia. Tree-ring-based streamflow reconstructions for the Sacramento River and Blue River (in the Upper Colorado River watershed) show concurrent drought conditions in both of these watersheds in the late-sixteenth century. This was one of the few periods of drought shared by both the Sacramento and the Blue River reconstructions in over 500 years and common to both records. During the pe- riod from 1580 to 1585, there were 4 years with concurrent drought conditions in both watersheds. Drought was particularly severe in the Sacramento River reconstruction, which indicated the driest 3-year period in the entire reconstruction (extending to AD 869) was 1578 to 1580. In addition to the western United States, there is also evidence of

312 ChapterFour

severe sustained drought in the western Great Plains about this time, with widespread mobilization of sand dunes in eastern Colorado and the Nebraska Sand Hills.

Analysis of Ni et al. (2002) is one of the cases studies confirming the late-sixteenth cen- tury megadrought and indicating similarly severe earlier droughts in Arizona and New Mexico. The authors developed a 1000-year reconstruction of cool-season (November– April) precipitation for each climate division in Arizona and New Mexico from a network of 19 tree-ring chronologies in the southwestern USA. Linear regression (LR) and artificial neural network (NN) models were used to identify the cool-season precipitation signal in tree rings. By using 1931 to 1988 records, the stepwise LR model was cross-validated with a leave-one-out procedure and the NN was validated with a bootstrap technique. The final models were also independently validated using the 1896 to 1930 precipitation data. In most of the climate divisions, both techniques can successfully reconstruct dry and normal years, and the NN seems to capture large precipitation events and more vari- ability better than the LR. In the 1000-year reconstructions, the NN also produces more distinctive wet events and more variability, whereas the LR produces more distinctive dry events. Figure 4.11 shows the results of the combined model (LR + NN) for one of Arizona’s climate divisions.

80 (mm) n o ti 60 ita ip 40

ip 40 Prec

F IGURE 4.11 A 1000-year reconstruction of cool-season (November–April) precipitation for the Arizona Climate Division No. 5. Bottom: 21-year moving average. (Source of raw data: Ni et al., 2002.)

ClimateChange

Several observations can be made from the bottom graph, which depicts the 21- year moving average. Among sustained dry periods comparable in duration and low precipitation to the observed 1950s drought, only two have somewhat similar wet periods preceding the droughts: 1730s and mid-1600s. In both cases, however, the sustained low precipitation is lower than the 1950s drought, and the preceding wet periods generally have lower precipitation. The fifteenth century megadrought occurred after a long period of average precipitation, without any significant wet periods. Even worse conditions are visible for the late-1200s megadrought, which was preceded by two shorter droughts during the first half of the thirteenth century. The droughts of late 1000s and mid-1100s were also likely more severe than the 1950s drought. Finally, the top graph with the model result on the annual basis shows that there were four seasons without any precipitation, the situation not recorded for the observed 1896 to 1988 period.

In the most recent study of paleo flows in the Colorado River Basin based on tree- ring records, Meko et al. (2007) show a very good agreement between the mid-1100s precipitation drought in central Arizona apparent in Fig. 4.11 and the Colorado River flow at Lee Ferry. The corresponding hydrologic drought is the most extreme low-frequency feature of the new reconstruction, covering AD 762 to 2005. It is characterized by a decrease of more than 15 percent in mean annual flow averaged over 25 years and by the absence of high annual flows over a longer period of about six decades. The drought is consistent in timing, with dry conditions inferred from tree-ring data in the Great Basin and Colorado Plateau, but regional differences in intensity emphasize the importance of basin-specific paleoclimatic data in quantifying likely effects of drought on water supply (Meko et al., 2007).

The National Climatic Data Center of the NOAA has compiled a Web site of exist- ing hydroclimatic reconstructions (streamflow, precipitation, and drought indices) for California based on tree-ring data. The site also shows locations of existing tree-ring chronologies that could be used to generate additional reconstructions. Links are pro- vided for similar information for the Colorado River Basin, an important source of wa- ter supply for southern California (http://www.ncdc.noaa.gov/paleo/streamflow/ca/ reconstructions.html).